Aging is associated with major changes in the structure and quality of sleep. Older subjects experience decreased sleep efficiency, increased numbers of arousals and awakenings, and sleep fragmentation, which may well play a pivotal role in both the daytime sleepiness and the age-associated decline in neurocognitive function in the elderly. This application develops and tests an integrative hypothesis that relates arousals to the occurrence of transient EEG events indicative of subthreshold arousals and sleep consolidations. By applying this integrative framework to both middle-aged (40-49 yrs) and elderly (70-79 yrs) Caucasian and African-American men and women, we expect to generate novel insights into the causes of sleep fragmentation in the elderly. Preliminary results show that Sample Entropy of the C3A2 EEG signal varies systematically with sleep state, being highest in Wake and lowest in stage 3-4, and that this variation is age-dependent. Thus, a conceptual model of temporal fluctuations of depth of sleep within a sleep state and a novel measure of depth of sleep, EEG entropy (a measure of irregularity), are proposed. A core principle of the model is that depth of sleep is continuously graded. Three hypotheses derived from this model will be tested: (1) EEG entropy fluctuates in a given sleep state and arousals are more likely when entropy transiently increases; (2) in a given sleep state, fluctuations of EEG entropy with time (which reflect short-lived events in the EEG such as microarousals and bursts of delta waves) are not entirely random; rather, they also reflect the slow decay of the effects of prior events that perturbed the depth of sleep; and, (3) the time course of decay of the effects of past sleep-disturbing events is slower in elderly than in middle- aged subjects; this factor contributes to the increased number of arousals in the elderly by prolonging the period of increased susceptibility to a subsequent sleep disturbance. This Small Research Grant will utilize existing data from the NHLBI-supported Sleep Heart Health Study (polysomnograms from >6400 subjects are available). PUBLIC HEALTH RELEVANCE: We expect that the proposed studies will develop and validate a novel framework for understanding the time-varying relationship of transient EEG events to arousals and their increase with aging. Additionally, Sample Entropy of EEG signals may find utility as a measure of depth of sleep for evaluating the effects of therapies. The goal of this study is to understand why the elderly awaken more frequently at night than younger individuals. To accomplish this objective, we will analyze electroencephalogram (EEG) recordings obtained from middle-aged and elderly subjects while they are asleep to enhance our understanding of the changes that occur in these EEGs immediately before subjects wake up. [unreadable] [unreadable] [unreadable] [unreadable]